Which deconvolution tool should you trust?
The problem. Spot-based assays like Visium capture several cells per spot, so nearly every downstream question depends on deconvolution — estimating the cell-type mixture inside each spot from a single-cell reference. There are a dozen methods and strong opinions; this paper asks which actually hold up, and when.
The idea. A systematic benchmark of spatial-deconvolution methods across simulated and real datasets, scoring accuracy against known or reference-derived mixtures and varying the conditions that matter — reference quality, number of cell types, spot resolution, and platform. The recurring result behind the field’s folk wisdom is that a handful of methods (notably cell2location, RCTD, and CARD) tend to lead, but the more useful contribution is the practical guidelines: which method to reach for given your reference, resolution, and tissue.
Why it matters. Deconvolution is the single most load-bearing step for spot assays, so a standards-minded facility prizes exactly this kind of method-evaluation — not “here’s a new tool” but “here’s which existing tool to trust, and under what conditions.” Being able to point to a benchmark rather than a preference is the difference between an opinion and a recommendation, and it’s the same evaluation discipline I’d apply to any pipeline component.
Verdict. The strongest kind of paper for practical work: it evaluates rather than advocates, and gives conditional guidance instead of a single winner. The honest caveat is that benchmark rankings depend on the simulation and reference choices, and imaging assays (single-cell resolution) shift the problem from deconvolution to segmentation. Read it as the decision guide for Visium-style data, paired with the cell2location and RCTD method papers.